Shaping can be very useful in training our dogs: it lets the dogs think for themselves, and make an active choice to perform. Their hunting instinct is activated, making learning through shaping fun for even breeds less eager to please! For example terriers or hunting dogs like beagles are happier to work independently from their handlers. Shaping lets the dog make the call.
One definition of shaping is the reinforcement of incremental, cumulative steps that that make up our desired behaviour. This technique can be used in combination with luring and is also related to capturing a behaviour.
As dogs get used to learning through shaping, they dare to take bigger steps towards what might be our desired behaviour: they get braver about offering us things. Especially working with props becomes very fun and interesting for the dog, as any prop can become the start of a fun training session! So while shaping can really be used for many things, it is especially interesting for training motivation, commitment and confidence! Shaping can also be used with zoo animals for instance. If people can train tigers to give blood samples via shaping, while standing on the other side of a fence -then you can definitely shape your dog no problem ????
To help our dogs understand the process, we need to set them up for success. To feel confident and bold enough to try things, they need to experience far more successful attempts than failures. This means that we will break down the behaviour we want to see into several simple chunks. Each chunk must be sufficiently reinforced before we raise our criteria and move on to the next step: we’ll always be expecting our dog to do a bit more work to receive his reward.
For a complete beginner, we’ll first reward ANY behaviour the dog offers!! We want the dog to know that he’s in control of what happens, and that daring to try things is good and fun. So if he chooses to take part in your session and offers any kind of behaviour – even just looking at you or sitting down: click – treat!
Now, to break down the behaviour we want to see, we need to think of every single mini piece of movement that leads towards what we want our dog to do. You can also consider this yourself: make a list of every single tiny movement you do as you go from being sat down to stood up. Whilst this is one position change, it involves many little limb movements as you adjust their position and weight distribution. If you were shaping someone to make this sitting to standing position change, it’d be best to reinforce each step on your long list!
It’s best to work on the small steps properly before asking for more of our desired behaviour. As with many things, it’s important to reinforce solid foundations before trying to build on them – otherwise you end up with an unstable behaviour. Just like trying to build a house on weak foundations is a poor idea, so is building a behaviour without the basics!
In the description, there are twenty steps that lead towards teaching your dog a paw target. Don’t forget to place your desired target in between your dog and you to start with, so the object is obviously in your space. I recommend using a larger target to start with, again to set your dog up for success. Make sure the target is safe! Nothing sharp or slippery.
- Dog is interested
- Dog is in the proximate area of yourself/target if the target is further away
- Dog looks towards target
- Dog looks at target
- Dog looks at target + duration
- Dog leans towards target
- Dog shifts weight off of one front paw
- Dog lifts paw
- Dog lifts paw off ground
- Paw moves towards target
- + each step towards target
(dog is now within touching distance of target)
- Dog’s head moves down towards target
- Dog’s nose touches target
- Dog lifts weight off one front paw
- Paw lifts
- Paw goes towards target
- Paw touches target
- Paw touches target + duration
- Paw touches target and bears weight
Using the rewards
To help your dog better understand, we can help with how and where we use our “primary reinforcer”. This is what the dog’s really working for: generally food or a toy, although of course our praise and presence are also rewarding to our dogs. Put some energy into thinking what your dog wants – on that day.
Our reward placement really impacts the learning process! Ideally, we need to reward most often where we want our dog to concentrate, otherwise he might not perform as intensely as we would like. If we want him to lie down, the food should be on the ground between his paws. If we want him to target with his nose, the food should be on the target. If we want him to wrap a jump wing, the reward should be at the base of the wing. Always consider where you want the value of the behaviour to be.
For instance, if I’ve got a small dog and I want him to focus on something low to the ground – like a lying down position, I could help him to succeed faster my sitting on the ground myself. Then his focus is already in the correct area. Equally, if I’m shaping a larger dog to stand on his back legs for example, I could help out by standing up.
You can even use a jackpot for really great suggestions. A larger amount, or higher value of reward can help highlight what the dog just did, and let him know that he made a great decision: for example if he offers a larger chunk of behaviour than expected.
Rewards can however also be useful for resetting the dog, so don’t hesitate to reward him out of your training zone. This can allow him to choose whether to return to training, but also help to re-approach the session with a slightly fresher mind. In some exercises this can mean for example rewarding the desired behaviour, then throwing one treat a bit away from your active zone before doing another repetition.
Ok, so so can I move from one criteria step to another?
Can your dog consistently offer what you’ve been working on? If yes, then let’s go!
Now, we will now reward next time the dog offers step x. If he’s learnt well enough that this should have earned him something good, he’ll think it’s a bit odd that nothing happened this time! This tiny “huh, that’s odd” feeling should drive your dog to try harder next time: he might look again at the object, move a bit closer, touch it a bit harder or longer.
This behavioural phenomenon can be called an extinction burst. When something we know works, stops working, we’ll try again HARDER! Until either it works again, and we’ve learnt that this item needs a bit more force – or it doesn’t work anymore, and our attempts will increase before they die out: the behaviour will (for this session) be extinct. There are so many examples this can be applied to, like trying to start our car when it’s broken: we’ll turn the key once. Huh? It doesn’t turn on. We’ll turn the key again, still nothing. Pull the key out, look at it, look at the ignition. Try again, harder. Now we’re a bit grumpy because it’s still not working! If someone was training us to just turn our key harder, they waited too long!
As a trainer, you want this transition to be as fluid as possible, to avoid the dog feeling helpless, stressed and frustrated. You want to observe your dog carefully, ride the wave of the behaviour burst and not ask for too much of an increase! If what you want next is too different from the previous step, it’ll be difficult to obtain without frustrating the dog. We want to keep the success rate very high, so the dog’s motivation doesn’t die out.
On the other hand, to prevent frustration of a different kind: don’t not move. When we say your dog consistently offers the behaviour, we don’t mean repeat a hundred times until perfect. If you haven’t yet reached your final criteria, don’t wait until the dog performs perfectly to raise the bar. Move to the next step when you get about three or four out of five tries correctly! If you reward each step until it becomes fluent, you’re actually making your dog work harder: every time to move forward, he would have to shut down the current behaviour. Shaping is exactly what the word would suggest – shaping until your end goal, not climbing big steps and stopping between each. When you’re where you want to be, you can keep repeating and rewarding the same behaviour to build confidence, speed and precision.
You can have fun, too!
All this sounds super serious right? But please don’t forget to enjoy it yourself, too! Admire your dog and all the great things he tries. Don’t have super high expectations, and make sure you’re praising him, smiling and interacting with him. So many studies have shown that dogs can read and understand our body language and verbal sounds, so he’ll know if you’re feeling happy!
My very own thesis study investigated the effect of handler behaviour on dog learning to target via shaping. It showed that dogs offer their taught behaviour faster if their handler was interactive and encouraging during the training session, an effect that was even bigger in beginner dogs who were not used to shaping or the clicker. Plus, there was no difference in results depending on age, breed type or working experience.
So, go and have a fun time with your dog, shaping is super interesting and really awesome to use as a training technique. Just remember to keep your sessions short and interesting, increase criteria gradually, and keep calm and happy! ????
Writer has got an MSc in applied animal behaviour and training, and works as a guide dog trainer in Belgium.